StructRSMA: A Proof-of-Concept Framework for PDB-Derived Contact-Supervised Transfer in RNA-Small Molecule Binding Affinity Prediction
- 1National University of Singapore , , ,
- 2Taipei Medical University , , ,
- 3Taipei Medical University Hospital , , ,
Abstract
RNA-small molecule affinity data sets usually provide only pair-level labels, making it difficult for multiview models to learn local nucleotide-atom recognition patterns. We present StructRSMA, a proof-of-concept contact-supervised transfer framework built on the DeepRSMA backbone. StructRSMA preserves the RNA sequence, RNA graph, molecule sequence, molecule graph, cross-fusion, and base affinity components, while introducing PDB-derived nucleotide-atom contact supervision and a Structural Contact Adapter (SCA). In Stage 1, RNA-ligand complexes are converted into binary contact maps using a 4 Å cutoff, and the shared backbone and contact head are trained to rank local contacts. In Stage 2, the contact-pretrained parameters are transferred to R-SIM, where only pKd labels are used; the frozen contact head supplies inferred contact priors to SCA for residual affinity calibration. Under a fixed validation-selected independent-test protocol, the reproduced DeepRSMA baseline obtained PCC 0.387 ± 0.022 and RMSE 0.970 ± 0.012. Contact500 pretraining without SCA yielded PCC 0.418 ± 0.048 and RMSE 0.951 ± 0.023, whereas full StructRSMA reached PCC 0.441 ± 0.037 and RMSE 0.918 ± 0.020. No-pretraining, shuffled-contact, and parameter-matched adapter controls did not reproduce the full improvement. Feature-removal and view-randomization analyses further indicated that contact statistics and multiview context contribute jointly to SCA. Contact prediction showed probability enrichment above contact density, but weak thresholded classification, and improvements became smaller and statistically uncertain under deoverlapped and cold evaluations. StructRSMA therefore provides evidence that PDB-derived contact supervision can support affinity calibration under covered data distributions, while explicitly identifying the limits of structural transfer.
Related
- A Single Active-Site Water Molecule Encodes Two Distinct Proton-Coupled Electron Transfer Pathways in Fatty Acid Photodecarboxylase
- Steering interfacial molecular structures by substrate-molecule charge transfer.
- Charge-transfer dynamics in bowl-shaped imide-functionalized corannulenes at ensemble and single-molecule levels
- Simulation-Supervised Foundation Models for Retention Time Prediction in High-Performance Liquid Chromatography beyond Experimental Data Coverage
- Times are changing but order matters: transferable prediction of small-molecule liquid chromatography retention times
BibTeX
@article{wen2026structrsma,
title = {{StructRSMA: A Proof-of-Concept Framework for PDB-Derived Contact-Supervised Transfer in RNA-Small Molecule Binding Affinity Prediction}},
author = {Yuxin Wen and Yi Tian and Nguyen Quoc Khanh Le and Matthew Chin Heng Chua},
journal = {ACS Omega},
year = {2026},
month = nov,
doi = {10.1021/acsomega.6c09001},
url = {https://pubs.acs.org/acsodf/article/doi/10.1021/acsomega.6c09001/5466094/StructRSMA-A-Proof-of-Concept-Framework-for-PDB},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - StructRSMA: A Proof-of-Concept Framework for PDB-Derived Contact-Supervised Transfer in RNA-Small Molecule Binding Affinity Prediction
AU - Yuxin Wen
AU - Yi Tian
AU - Nguyen Quoc Khanh Le
AU - Matthew Chin Heng Chua
PY - 2026
DA - 2026/11/06
JO - ACS Omega
DO - 10.1021/acsomega.6c09001
UR - https://pubs.acs.org/acsodf/article/doi/10.1021/acsomega.6c09001/5466094/StructRSMA-A-Proof-of-Concept-Framework-for-PDB
AB - RNA-small molecule affinity data sets usually provide only pair-level labels, making it difficult for multiview models to learn local nucleotide-atom recognition patterns. We present StructRSMA, a proof-of-concept contact-supervised transfer framework built on the DeepRSMA backbone. StructRSMA preserves the RNA sequence, RNA graph, molecule sequence, molecule graph, cross-fusion, and base affinity components, while introducing PDB-derived nucleotide-atom contact supervision and a Structural Contact Adapter (SCA). In Stage 1, RNA-ligand complexes are converted into binary contact maps using a 4 Å cutoff, and the shared backbone and contact head are trained to rank local contacts. In Stage 2, the contact-pretrained parameters are transferred to R-SIM, where only pKd labels are used; the frozen contact head supplies inferred contact priors to SCA for residual affinity calibration. Under a fixed validation-selected independent-test protocol, the reproduced DeepRSMA baseline obtained PCC 0.387 ± 0.022 and RMSE 0.970 ± 0.012. Contact500 pretraining without SCA yielded PCC 0.418 ± 0.048 and RMSE 0.951 ± 0.023, whereas full StructRSMA reached PCC 0.441 ± 0.037 and RMSE 0.918 ± 0.020. No-pretraining, shuffled-contact, and parameter-matched adapter controls did not reproduce the full improvement. Feature-removal and view-randomization analyses further indicated that contact statistics and multiview context contribute jointly to SCA. Contact prediction showed probability enrichment above contact density, but weak thresholded classification, and improvements became smaller and statistically uncertain under deoverlapped and cold evaluations. StructRSMA therefore provides evidence that PDB-derived contact supervision can support affinity calibration under covered data distributions, while explicitly identifying the limits of structural transfer.
SN - 2470-1343
ER -